Skip to content

[Bioc-devel] Query about microshades package suitability for Bioconductor

4 messages · Scott McLoud, Lluís Revilla

Scott McLoud
# 
Hello,

I'm a graduate student working on preparing the R package "microshades" for submission to BioConductor.

Microshades is an R package  designed to provide custom color shading palettes that improve accessibility and data organization.

My question is, does this fulfill BioConductor's requirements of addressing areas of high-throughput genomic analysis?

The GitHub page for the R package can be found here: https://github.com/KarstensLab/microshades

Scott
Lluís Revilla
# 
Dear Scott,

Sorry for not answering the first time you asked.
I thought someone else would give you some feedback but I guess no one did
it off-list.

How is this specific palette different from other palettes existing on CRAN
or Bioconductor?
What makes this specific palette better for Bioconductor packages (as
opposed to CRAN packages)? How does the repository or the package benefit
from being together with the other cohort of packages?

I think as a color palette is quite general purpose and doesn't require
much maintenance, your package might be better suited for CRAN. There you
can release one and if the checks pass you won't be required to update it
anymore.

Best wishes,

Llu?s Revilla

On Tue, 20 May 2025 at 18:20, Scott McLoud via Bioc-devel <
bioc-devel at r-project.org> wrote:

            

      
      
      
    

        

      
      
    

        

  
  


  


      

    

    
7 days later

    

      
      

        


  

        

      Scott McLoud
    

    

      
      #
    

  


  

      
Hello Lluis,


Dear Llu?s,



How is this specific palette different from other palettes existing on CRAN or Bioconductor? This color palette is different than others on CRAN or Bioconductor in that it is meant to help users with visualizing complex hierarchical data which is common in biology (specifically microbiome sequencing data) while also retaining accessibility to individuals with color vision deficiencies / colorblindness, as described in the microshades manuscript<https://journals.asm.org/doi/10.1128/mra.00795-22>. In addition to the color palette, it has several functions that help users organize and visualize their data.



What makes this specific palette better for Bioconductor packages (as opposed to CRAN packages)? How does the repository or the package benefit from being together with the other cohort of packages? The package was specifically written to work with phyloseq, which is a Bioconductor package that supports microbiome data analysis  and uses a specialized system of S4 classes to store all related phylogenetic sequencing data as single experiment-level object, making it easier to share data and reproduce analyses.   We have also received requests to extend microshades to other Bioconductor packages (TreeSummarizedExperiment) and to release it through Bioconductor so that it can be used as a dependency in other Bioconductor packages.



I think as a color palette is quite general purpose and doesn't require much maintenance, your package might be better suited for CRAN. There you can release one and if the checks pass you won't be required to update it anymore. Since there is functionality in addition to the color palette itself, it may require regular maintenance as the packages it is meant to be used with are updated, and as we expend to additional applications (such as to work with the TreeSummarizedExperiment and other biological data with hierarchical structures).



Though if you feel this is not a good fit for Bioconductor, we will submit to Cran


Regards,

Scott McLoud

        

      
      
    

  
  


  


      

    

    
1 day later

    

      
      

        


  

        

      Lluís Revilla
    

    

      
      #
    

  


  

      
Dear Scott,

I'll reply within the content:

        
On Tue, 27 May 2025 at 21:12, Scott McLoud <mclouds at ohsu.edu> wrote:

        

            

      
      
      
    

        
I see that the value of the package is not on the color palettes but in the
combination of the palette and these methods.

            

      
      
      
    

        
Even if some developers prefer not to mix Bioconductor and CRAN packages,
you can extend Bioconductors packages while your package is available on
CRAN. Similarly, Bioconductor packages depend and extend those on CRAN.
Last time I checked, CRAN had 722 packages that used Bioconductor (~3%),
Bioconductor had 2802 packages that used CRAN's packages (~77%).

            

      
      
      
    

        
submit to Bioconductor.
Once your package passes the automatic checks, the repository reviewers,
after (carefully) analyzing the code, will decide if your package is
accepted and published wherever you submit it.

Good luck!

Llu?s